Provided here are a few python scripts for interacting with your own locally hosted GPT4All LLM model using Langchain. There is also a script for interacting with your cloud hosted LLM's using Cerebrium and Langchain The scripts increase in complexity and features, as follows:
local-llm.py
Interact with a local GPT4All model.
local-llm-chain.py
Interact with a local GPT4All model using Prompt Templates.
cloud-llm.py
Interact with a cloud hosted LLM model. The model is deployed and hosted on the Cerebrium platform.
I wrote an article which explores some of the concepts here, as well as walks through building each of the scripts. Can read that here
Clone the repository, set up the virtual environment, and install the required packages
git clone [email protected]:smaameri/private-llm.git
cd private-llm
mkdir models
python3 -m venv .venv
. .venv/bin/activate
pip install -r requirements.txt
To use the cloud-llm.py script that interacts with an LLM deployed on Cerebrium, you will need to setup your public Cerebrium API Key.
Copy the example env file
cp .env.example .env
Now copy your public Cerebrium API key into the .env
file, and save the file. It should send up looking something like
CEREBRIUMAI_API_KEY=public-
To start interacting with a local-llm, you can download the GPT4All LLM from here https://gpt4all.io/index.html, and
add the downloaded model to the projects /models
directory.
Now kick off the local-llm-chain.py script to start interacting with it.
python3 local-llm-py.py
Could it be better? Try tweaking the prompts to get a better chatbot style assistant. Interested in using a local LLM to read and interact with documents? Combine the concepts here with the project I created on building a document reader and chatbot.